1940453 - Crash in [@ mozilla::gl::ScopedSaveMultiTex::ScopedSaveMultiTex]

Status: RESOLVED FIXED

(Core :: Widget: Gtk, defect, P2)

Description

[Donal Meehan [:dmeehan]]

Crash report: https://crash-stats.mozilla.org/report/index/0de284be-f206-491f-8380-71e560250108

MOZ_CRASH Reason:

MOZ_RELEASE_ASSERT(texUnits >= 1)

Top 10 frames:

0  libxul.so  mozilla::gl::ScopedSaveMultiTex::ScopedSaveMultiTex(mozilla::gl::GLContext*, ...  gfx/gl/GLBlitHelper.cpp:339
1  libxul.so  mozilla::gl::GLBlitHelper::Blit(DMABufSurface*, mozilla::gfx::IntSizeTyped<mo...  gfx/gl/GLBlitHelper.cpp:1560
2  libxul.so  DMABufSurfaceYUV::ReadIntoBuffer(unsigned char*, int, mozilla::gfx::IntSizeTy...  widget/gtk/DMABufSurface.cpp:1768
3  libxul.so  DMABufSurfaceYUV::BuildSurfaceDescriptorBuffer(mozilla::layers::SurfaceDescri...  widget/gtk/DMABufSurface.cpp:1825
4  libxul.so  mozilla::RemoteDecoderManagerParent::RecvReadback(mozilla::layers::SurfaceDes...  dom/media/ipc/RemoteDecoderManagerParent.cpp:295
5  libxul.so  mozilla::PRemoteDecoderManagerParent::OnMessageReceived(IPC::Message const&, ...  ipc/ipdl/PRemoteDecoderManagerParent.cpp:404
6  libxul.so  mozilla::ipc::MessageChannel::DispatchSyncMessage(mozilla::ipc::ActorLifecycl...  ipc/glue/MessageChannel.cpp:1763
7  libxul.so  mozilla::ipc::MessageChannel::DispatchMessage(mozilla::ipc::ActorLifecyclePro...  ipc/glue/MessageChannel.cpp:1715
7  libxul.so  mozilla::ipc::MessageChannel::RunMessage(mozilla::ipc::ActorLifecycleProxy*, ...  ipc/glue/MessageChannel.cpp:1508
8  libxul.so  mozilla::ipc::MessageChannel::MessageTask::Run()  ipc/glue/MessageChannel.cpp:1608

Comment 1

[Donal Meehan [:dmeehan]]

:stransky looks like this might have been introduced by Bug 1940037?
Could you take a look?
(Waiting for confirmation before setting the regressor)

Comment 2

[Martin Stránský [:stransky] (ni? me)]

Sure, will look at it.

Comment 3

[Donal Meehan [:dmeehan]]

Thanks.
Once you've confirmed, I might ask the sheriffs to back out and trigger a new nightly Linux build.

Comment 4

[Martin Stránský [:stransky] (ni? me)]

This is not a regression caused by Bug 1940037. The patches there reveals underlying bug when we're blitting already released surface.
Before Bug Bug 1940037 was hard coded but now it return real plane count so it returns 0 for empty surfaces which should not be used.
Correct fix is to ensure we don't process empty surfaces.

Comment 5

[Martin Stránský [:stransky] (ni? me)]

DMABufSurfaceYUV::ReadIntoBuffer() needs to be fixed, will look at it.

Comment 6

[Donal Meehan [:dmeehan]]

Any objection if we back out Bug 1940037 in the meantime?

Comment 7

[Martin Stránský [:stransky] (ni? me)]

(In reply to Donal Meehan [:dmeehan] from comment #6)

Any objection if we back out Bug 1940037 in the meantime?

Please keep it there, will look at it ASAP.

Comment 8

[Martin Stránský [:stransky] (ni? me)]

Attachment: Bug 1940453 [Linux] Don't read empty DMABuf surfaces r?emilio

Comment 9

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/89b662e27321
[Linux] Don't read empty DMABuf surfaces r=emilio

Comment 10

[Serban Stanca [:SerbanS]]

https://hg.mozilla.org/mozilla-central/rev/89b662e27321